RE: [Idr] Re: Last Call: 'Canonical representation of 4-byte ASnumbers' to Informational RFC(draft-michaelson-4byte-as-representation)

["Bruno Rijsman" <brijsman at juniper.net>]

I second this, in other words: I am also in favor of simply displaying
4-octet AS numbers as (potentially large) decimal integer numbers.

There is already one widely used implementation that does just that.

The main reason for my preference is that there is no semantic
difference between 333 and 0.333: they both refer to the same autonomous
system. Consider an AS-path with traverses various islands of 2-octet
and 4-octet BGP speakers; it would probably be confusing if the AS
number flipped back and forth from 333 to 0.333.

-- Bruno

> -----Original Message-----
> From: Bill Fenner [mailto:fenner at research.att.com]
> Sent: Tuesday, October 10, 2006 9:37 AM
> To: Joe Abley
> Cc: idr at ietf.org; John Leslie; iesg at ietf.org
> Subject: Re: [Idr] Re: Last Call: 'Canonical representation of 4-byte
> ASnumbers' to Informational
> RFC(draft-michaelson-4byte-as-representation)
> 
> 
> 
> >[If you just use larger decimal values,]
> >... how do you distinguish between a 2-byte only AS number and a  
> >4-byte AS number which happens to be less than 0.65535?
> 
> When is this difference significant?  0.7018 == 7018, an AS number
> is an AS number.  There are some AS numbers that can *only* be
> represented in 4 bytes, but there isn't a different space for the
> 4-byte values.
> 
> I'm in the "having a common notation is better than not" camp 
> - there is
> no perfect notation, but if everyone ends up using the same notation,
> it'll improve consistency and reduce confusion.  (I'd be willing to be
> convinced "no notation is OK", i.e., people don't have a problem with
> using longer and longer decimal strings - we probably don't expect to
> get to 4294967295 but some would argue that's too long for people to
> easily remember)
> 
>   Bill
> 
> _______________________________________________
> Idr mailing list
> Idr at ietf.org
> https://www1.ietf.org/mailman/listinfo/idr
> 

_______________________________________________
Idr mailing list
Idr at ietf.org
https://www1.ietf.org/mailman/listinfo/idr